Evolutionary network analysis based on the transcriptome atlas of Marchantia polymorpha (EvoNet)

基因调控网络通过相互作用的转录因子控制所有基因在植物中的表达，从而最终控制植物的表型。该网络高度复杂，经常包含冗余元件，并且由于基因组复杂性的扩展，其复杂性通常从苔藓上的藻类扩展到种子植物。地钱为我们提供了一个了解早期陆地植物进化和陆地植物转录因子进化历史的独特窗口。Marchantia缺乏全基因组复制，因此保留了一个相当简单的调控电路，总共只有374个转录因子。这种简单性将有助于研究代谢途径、细胞分化和对内部和外部信号的反应的转录控制。我们将构建一个地钱的转录图谱，包括发育、胁迫和信号诱导的时间过程，目的是捕捉可能发生的大部分转录变化。有监督的机器学习方法利用这个转录图谱来预测转录因子与靶基因的相互作用，并构建一个基因调控网络。我们打算(I)为Marchantia群落和我们自己提供一个全面的、开放的、容易获得的基因表达资源，(Ii)利用这个资源来建立和分析地钱属植物的基因调控网络。我们将测试与地钱相比，亚功能化和新功能化在多大程度上增加了种子植物调控网络的复杂性。函数的实验验证将检验通过计算产生的假设。

Gene regulatory networks control the expression of all genes in plants via interacting transcription factors and therefore ultimately control the phenotypes plants display. The network is highly complex, frequently contains redundant elements, and generally expands in complexity from alga over mosses towards seed plants since genomic complexity expands. The liverwort Marchantia polymorpha presents us with a unique window into early land plant evolution and the evolutionary history of transcription factors in land plants. Marchantia lacks whole genome duplications and therefore retains a fairly simple regulatory circuitry with only 374 transcription factors in total. This simplicity will facilitate the study of transcriptional control for metabolic pathways, for cellular differentiation, and for responses to internal and external signals. We will construct a transcriptome atlas of the liverwort including developmental and stress and signal induced time courses which aims at capturing the majority of transcriptional changes that can occur. Supervised machine learning approaches make use of this transcriptional atlas to predict transcription factor-target gene interactions and construct a gene regulatory network. The presumably much simpler network of a liverwort can then be compared them with those in seed plant networks and the mechanisms for expansion studied.We intend to (i) provide a comprehensive, open, easily accessible gene expression resource for the Marchantia community and ourselves and (ii) leverage the resource to build and analyze gene regulatory networks in the liverwort. We will test to which degree sub- and neo-functionalization have contributed to increased complexity of regulatory networks in seed plants compared to liverworts. Experimental validation of function will test the hypotheses which were computationally generated.